Diet and hormonal status are relevant to the etiology of colon cancer. The role of estrogens in the etiology of the disease has been strengthened considerably by the discovery that virtually all colonic tumors arise from epithelial cells that show hypermethylation of the estrogen receptor (ER) gene. Foods high in plant estrogens, (particularly soy isoflavones) are associated with reduced risk of colon neoplasia in epidemiologic studies. We hypothesize that isoflavones will lower risk of colorectal neoplasia by slowing or reversing the degree of methylation of the ER gene and by reducing colonic epithelial cell proliferation. We plan to test the hypothesis in a randomized, double-blind trial of isoflavone-rich and isoflavone-poor soy supplements. We further hypothesize that specific ER/estrogen responsive genes that are relevant to colon cancer will show a more controlled pattern in the presence vs. absence of isoflavones. A double-blind, randomized trial involving 160 individuals - 80 men, and 80 women not using hormone replacement therapy (HRT) - will be conducted. Eligible individuals with a recent history of adenomatous polyps, aged 50- 74, will be randomized (blocking on sex) to an isoflavone-poor soy supplement over a period of 12 months. Colonic biopsies (proximal and distal) will be taken at baseline and 12 months, and examined for degree of ER gene methylation, epithelial cell proliferation, and expression of specific estrogen-responsive markers - connexins (gap junction proteins), E-cadherin (a cell adhesion molecule), and bcl-2 and bax (an inhibitor and an inducer of apoptosis, respectively). Dietary compliance will be monitored by serum isoflavone concentration. We postulate that men and women in the isoflavone-rich arm will show a more favorable pattern of markers at 12 months than those on the placebo. Results from this human experimental study will provide important data on mechanisms of carcinogenesis and on possible preventive strategies in both sexes.